#define nv50_dmaobj(p) container_of((p), struct nv50_dmaobj, base)
#include "user.h"
#include <core/client.h>
#include <core/gpuobj.h>
#include <subdev/fb.h>
#include <nvif/cl0002.h>
#include <nvif/unpack.h>
struct nv50_dmaobj {
struct nvkm_dmaobj base;
u32 flags0;
u32 flags5;
};
static int
nv50_dmaobj_bind(struct nvkm_dmaobj *base, struct nvkm_gpuobj *parent,
int align, struct nvkm_gpuobj **pgpuobj)
{
struct nv50_dmaobj *dmaobj = nv50_dmaobj(base);
struct nvkm_device *device = dmaobj->base.dma->engine.subdev.device;
int ret;
ret = nvkm_gpuobj_new(device, 24, align, false, parent, pgpuobj);
if (ret == 0) {
nvkm_kmap(*pgpuobj);
nvkm_wo32(*pgpuobj, 0x00, dmaobj->flags0);
nvkm_wo32(*pgpuobj, 0x04, lower_32_bits(dmaobj->base.limit));
nvkm_wo32(*pgpuobj, 0x08, lower_32_bits(dmaobj->base.start));
nvkm_wo32(*pgpuobj, 0x0c, upper_32_bits(dmaobj->base.limit) << 24 |
upper_32_bits(dmaobj->base.start));
nvkm_wo32(*pgpuobj, 0x10, 0x00000000);
nvkm_wo32(*pgpuobj, 0x14, dmaobj->flags5);
nvkm_done(*pgpuobj);
}
return ret;
}
static const struct nvkm_dmaobj_func
nv50_dmaobj_func = {
.bind = nv50_dmaobj_bind,
};
int
nv50_dmaobj_new(struct nvkm_dma *dma, const struct nvkm_oclass *oclass,
void *data, u32 size, struct nvkm_dmaobj **pdmaobj)
{
union {
struct nv50_dma_v0 v0;
} *args;
struct nvkm_object *parent = oclass->parent;
struct nv50_dmaobj *dmaobj;
u32 user, part, comp, kind;
int ret;
if (!(dmaobj = kzalloc(sizeof(*dmaobj), GFP_KERNEL)))
return -ENOMEM;
*pdmaobj = &dmaobj->base;
ret = nvkm_dmaobj_ctor(&nv50_dmaobj_func, dma, oclass,
&data, &size, &dmaobj->base);
if (ret)
return ret;
ret = -ENOSYS;
args = data;
nvif_ioctl(parent, "create nv50 dma size %d\n", size);
if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
nvif_ioctl(parent, "create nv50 dma vers %d priv %d part %d "
"comp %d kind %02x\n", args->v0.version,
args->v0.priv, args->v0.part, args->v0.comp,
args->v0.kind);
user = args->v0.priv;
part = args->v0.part;
comp = args->v0.comp;
kind = args->v0.kind;
} else
if (size == 0) {
if (dmaobj->base.target != NV_MEM_TARGET_VM) {
user = NV50_DMA_V0_PRIV_US;
part = NV50_DMA_V0_PART_256;
comp = NV50_DMA_V0_COMP_NONE;
kind = NV50_DMA_V0_KIND_PITCH;
} else {
user = NV50_DMA_V0_PRIV_VM;
part = NV50_DMA_V0_PART_VM;
comp = NV50_DMA_V0_COMP_VM;
kind = NV50_DMA_V0_KIND_VM;
}
} else
return ret;
if (user > 2 || part > 2 || comp > 3 || kind > 0x7f)
return -EINVAL;
dmaobj->flags0 = (comp << 29) | (kind << 22) | (user << 20) |
oclass->base.oclass;
dmaobj->flags5 = (part << 16);
switch (dmaobj->base.target) {
case NV_MEM_TARGET_VM:
dmaobj->flags0 |= 0x00000000;
break;
case NV_MEM_TARGET_VRAM:
dmaobj->flags0 |= 0x00010000;
break;
case NV_MEM_TARGET_PCI:
dmaobj->flags0 |= 0x00020000;
break;
case NV_MEM_TARGET_PCI_NOSNOOP:
dmaobj->flags0 |= 0x00030000;
break;
default:
return -EINVAL;
}
switch (dmaobj->base.access) {
case NV_MEM_ACCESS_VM:
break;
case NV_MEM_ACCESS_RO:
dmaobj->flags0 |= 0x00040000;
break;
case NV_MEM_ACCESS_WO:
case NV_MEM_ACCESS_RW:
dmaobj->flags0 |= 0x00080000;
break;
default:
return -EINVAL;
}
return 0;
}